Research On Context Modeling And Reasoning In Ubiquitous Learning Environment

With the development of computer, mobile computing and sensor network, computing is becoming ubiquitous. The concept of ubiquitous computing is leading a new area for computing. The ubiquitous computing technology integrates the information space and physical space together, where the information space is constituted by communication and computer while the physical space serves for people living and working. In the ubiquitous computing environment, user is centered and is able to obtain personalized services at anytime and anywhere. The ubiquitous computing is nowadays applying in a lot of fields and ubiquitous learning is one of them. With the support of ubiquitous computing, each student can interact with multiple embedded equipments in the ubiquitous learning environment, according to the environment information that encompassing user; system will determine user's state for accurate and personalized services. As one of its intrinsic attributes, context-aware is the key technology to ubiquitous learning. The master Dey pointed out that context awareness is about capturing a broad range of contextual attributes (such as the user's current positions, activities, and their surrounding environments) to better understand what the user is trying to accomplish, and what services the user might be interested. Context awareness has the potential to greatly reduce the human attention and interaction bottlenecks, to give the user the impression that services fade into the background, and to support intelligent personalization and adaptability features. In this paper, we study the existing context modeling and reasoning technology and proposed activity-centric context model and reasoning algorithm in ubiquitous learning environment.Ubiquitous learning happens in ubiquitous learning environment, we analyze and categorize various types of contextual information at first and apply an extended ontology method to modeling and representing contexts for the purpose of time sequence of activity and knowledge sharing and reuse. Our activity-centric context model consists two parts, one is the process model which defines the sequence and conditions of activities, and the other part is the ontologies describe and interpret the specific contextual information associated with the activity. Context reasoning refers to the process that the system makes use of the contextual information about users and environment in order to acquire and extract high-level semantics. In ubiquitous learning environment, activity is recognized as the high-level context which is derived from the low-level, explicit contexts which are directly provided by hardware sensors and software programs and is on the basis of service provided by the system. According to the time sequence of context acquisition, we defined the context stream first. Since the activity context cannot be recognized directly be the sensor data, context pattern and context evolving pattern is defined to describe the activity. A context reasoning algorithm is proposed to fulfill the context reasoning from context stream to context evolving patterns. At last, we take the single-crystal X-ray diffraction experiment as an example to build formal context model and check the reasoning algorithm through several simulation scene.